A Dual Three-Phase Motor Model Predictive Current Control Method with Fixed Switching Frequency

Abstract

The present invention provides a method for model predictive current control of a dual three-phase motor with a fixed switching frequency, which relates to the technical field of dual three-phase permanent magnet synchronous motor control. First, a virtual voltage vector is synthesized by two large voltage vectors and two medium and small voltage vectors, reducing current harmonics, and the virtual voltage vector meets the conditions for generating a standard PWM wave. Secondly, a cost function without a harmonic term weight coefficient is designed. In addition, a selection process for the optimal voltage vector is designed to reduce the candidate voltage vectors and the computational burden of the controller. Finally, in order to further reduce the harmonic current, a duty cycle modulation technique is introduced, the dwell time of the effective vector is calculated, and the zero vector is directly arranged in the middle and on both sides of the sampling period, fixing the switching frequency without the need for secondary modification of the switching sequence, reducing the design difficulty.

Description

Technical Field

[0001] The present invention relates to the technical field of dual-three-phase permanent magnet synchronous motor control, and particularly to a model predictive current control method for a dual-three-phase motor with a fixed switching frequency. Background Art

[0002] The model predictive control method is considered to be one of the most effective control methods for multiphase drive systems due to its fast response ability, simple structure, and feasible multi-objective optimization. The main principle of the model predictive technology is to predict the current of all voltage vectors, and select the optimal voltage vector by comparing the expected designed cost function. The optimal voltage vector acts on the drive system at the next moment. The model predictive technology is widely applied to three-phase drive systems. In contrast, for multiphase systems, due to the existence of harmonic planes, if the harmonic voltages are not processed, a large amount of harmonic current will be generated, resulting in a large stator copper loss. Moreover, for multiphase systems, when designing the cost function, it is also necessary to consider introducing the weight coefficient of the harmonic term. Since the design of the weight coefficient has no theoretical basis support, most of them need to determine the parameters in practical applications, which brings a cumbersome task.

[0003] As the number of phases of the motor increases, the number of voltage vectors also increases exponentially. If all voltage vectors are traversed and calculated in the same way as in three-phase systems, it will bring a serious computational burden to the controller. Now, the methods for reducing voltage vectors all use the deadbeat model predictive voltage method to reduce the candidate voltage vectors. This method selects the candidate voltage vectors by calculating the sector position of the reference voltage vector. However, the calculation of the reference voltage vector involves complex coordinate transformations and trigonometric function calculations, etc., which are not desirable in the technology of automatic code generation. In addition, for the application of model predictive technology to multiphase motors, there is also the problem of non-fixed switching frequency. In actual industrial applications, a constant switching frequency is necessary because the application fields of multiphase motors are generally high-power occasions. If the switching frequency is large, it will cause excessive switching losses, resulting in an increase in the temperature of the switching devices, and factors such as heat dissipation need to be considered more. More importantly, in actual engineering, a constant switching frequency makes it easier to implement in software design and other aspects. Some scholars have also conducted detailed research on the fixed switching frequency problem of dual-three-phase permanent magnet synchronous motors, but the proposed methods all perform secondary correction on the PWM switching sequence, which increases the complexity of the algorithm to a certain extent. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides a model predictive current control method for a dual-three-phase motor with a fixed switching frequency.

[0005] A model predictive current control method for a dual-three-phase motor with a fixed switching frequency specifically includes the following steps:

[0006] Step 1: According to the space decoupling matrix, map the 64 switching states of the six-phase voltage source inverter to the α-β space containing electromechanical energy conversion and the x-y space with only harmonic components, and obtain the voltage vector distribution of the dual three-phase permanent magnet synchronous motor;

[0007] The voltage vector distribution, that is, the voltage distributions in the α-β space and the x-y space are shown as follows:

[0008]

[0009] where a = e j30° ; s represents the switching function of the inverter, s i = 1 represents that the upper bridge arm is conducting and the lower bridge arm is off, s i = 0 represents that the upper bridge arm is off and the lower bridge arm is conducting, i represents the A, B, C, U, V, W phases of the inverter; U dc represents the DC bus voltage of the inverter; V αβ and Vxy are the amplitudes in the α-β space and the x-y space; the six-phase voltage inverter generates 64 voltage vectors, including 60 effective voltage vectors and 4 zero vectors. The 64 basic voltage vectors are divided into four groups according to different amplitudes: large voltage vectors, medium-large voltage vectors, medium-small voltage vectors, and small voltage vectors, and the amplitudes are: 0.644U dc , 0.471U dc , 0.331U dc and 0.172U dc ;

[0010] Step 2: Discretize the mathematical model of the dual three-phase permanent magnet synchronous motor through the forward Euler formula to obtain the prediction model of the dual three-phase permanent magnet synchronous motor;

[0011] The mathematical model of the dual three-phase permanent magnet synchronous motor is shown as follows:

[0012]

[0013]

[0014] where u d and u q are the voltages on the d and q axes; i d and i q are the currents on the d and q axes; u x and u y are the voltages on the x and y axes; i x and i y are the currents on the x and y axes; R s is the stator resistance; ω r is the electrical angular velocity; Ld and L q are the inductances on the d- and q-axes; L ls represents the leakage inductance; ψ f represents the permanent magnet flux linkage;

[0015] Derivation is carried out using the forward Euler method:

[0016]

[0017] where x is a variable; T s represents the sampling period, k represents the k-th sampling period, k + 1 represents the (k + 1)-th sampling period, and the discrete expression of the dual three-phase permanent magnet synchronous motor is expressed as:

[0018]

[0019] where u d (k) and u q (k) are the voltages on the d- and q-axes at the k-th moment; i d (k) and i q (k) are the currents on the d- and q-axes at the k-th moment;

[0020] The current predictions on the d-axis and q-axis at the (k + 1)-th moment are:

[0021]

[0022] Using two-step prediction to compensate for the calculation delay, the current prediction at the instantaneous (k + 2) is:

[0023]

[0024] Step 3: Select two large vectors and two medium-small vectors to synthesize virtual voltage vectors, and synthesize virtual voltage vectors in sequence according to the duty cycles 0.432, 0.432, 0.068, 0.068, and a total of 12 virtual voltage vectors are synthesized;

[0025] The voltage amplitude of the virtual voltage vector in the x-y subspace is zero, and the expression is as follows:

[0026]

[0027] In the formula, 0.173U dc , 0.333U dc are the voltage amplitudes of the large voltage vector and the medium-small voltage vector in the x-y space respectively, where η1, η2, η3, and η4 are the duty cycles of the four voltage vectors V 04 , V 44 , V 45 , V 75 respectively, where:

[0028] η1 + η2 + η3 + η4 = 1 (9)

[0029] Substitute formula (9) into formula (8) to obtain the duty ratio as:

[0030]

[0031] In the α-β space, the amplitude of the virtual voltage vector is obtained according to the volt-second balance principle:

[0032]

[0033] |VV1| αβ is the amplitude of the virtual voltage vector in the α-β space;

[0034] Step 4: Design a value function G without weight coefficients as follows:

[0035]

[0036] where, i d * (k) and i q * (k) are the given values of the d-axis and q-axis currents respectively; i d (k + 2) and i q (k + 2) are the predicted estimated values of the currents at the (k + 2)-th moment on the d-axis and q-axis respectively.

[0037] Step 5: Design a selection process for the optimal voltage vector to reduce the candidate voltage vectors.

[0038] Step 5.1: First, perform predictive calculations on the virtual voltage vectors VV4, VV8, VV 12 to obtain the value functions of the three, namely G(VV4), G(VV8), and G(VV 12 ), compare the values of the three value functions, and select the one with the largest value function among the three and denote it as G max , and denote the minimum value function as G min ;

[0039] Step 5.2: Define the voltage vector with the largest value function G max as the worst voltage vector. The reference voltage vector is within a 1 / 3 circle range in the opposite direction of the worst voltage vector. G max determines the range of the reference voltage vector, and there is a voltage vector with a value function of G min at both ends of the range. Compare the value functions of the voltage vectors within the 1 / 6 circle range, and the one with the minimum value function is the optimal voltage vector.

[0040] Step 6: In each control period, the dwell time of the optimal voltage vector is determined by the duty cycle modulation technique;

[0041] The duty cycle of the optimal voltage vector is:

[0042]

[0043]

[0044] If d is the duty cycle of the optimal voltage vector, then the duty cycle of the zero vector is (1 - d)T s , T s is the control period.

[0045] The meanings of m and n in Equation (14) are as follows:

[0046]

[0047]

[0048] Step 7: A switching sequence with a fixed switching frequency is designed. Zero vectors V 00 and V 77 are inserted on both sides and in the middle of the switching sequence of the virtual voltage vector, that is, the action order of the voltage vectors, to achieve the purpose of a fixed switching frequency.

[0049] Step 7.1: The virtual voltage vector is synthesized from four voltage vectors, namely two large voltage vectors and medium and small voltage vectors. If the optimal voltage vector is one of the virtual voltage vectors VV1, VV4, VV5, VV8, VV9, and VV 12 If it is one of them, the switching sequence is designed such that in the first half of the control period, the four voltage vectors act in a clockwise direction in sequence, and in the second half of the period, the four voltage vectors act in a counterclockwise direction in sequence. If the optimal voltage vector is one of the virtual voltage vectors VV2, VV3, VV6, VV7, VV 10 , VV 11 If it is one of them, the switching sequence is designed such that in the first half of the control period, the four voltage vectors act in a counterclockwise direction in sequence, and in the second half of the period, the four voltage vectors act in a clockwise direction in sequence.

[0050] Step 7.2: After obtaining the switching sequence of the virtual voltage vector, zero vectors V 00 and V 77 are inserted on both sides and in the middle of the switching sequence of the virtual voltage vector respectively. After inserting the zero vectors, the six-phase switching devices are turned on and off once in one control period, achieving a fixed switching frequency and forming a standard PWM wave;

[0051] The switching sequence with the fixed switching frequency is shown in Table 1;

[0052] Table 1 Switch Sequence Table

[0053] Optimal voltage vector Switching sequence with fixed switching frequency <![CDATA[VV1]]> <![CDATA[V 00 →V 04 →V 44 →V 45 →V 75 →V 77 →V 77 →V 75 →V 45 →V 44 →V 04 →V 00 > <![CDATA[VV2]]> <![CDATA[V 00 →V 40 →V 44 →V 64 →V 67 →V 77 →V 77 →V 67 →V 64 →V 44 →V 40 →V 00 > <![CDATA[VV3]]> <![CDATA[V 00 →V 04 →V 64 →V 66 →V 76 →V 77 →V 77 →V 76 →V 66 →V 64 →V 04 →V 00 > <![CDATA[VV4]]> <![CDATA[V 00 →V 20 →V 26 →V 66 →V 67 →V 77 →V 77 →V 67 →V 66 →V 26 →V 20 →V 00 > <![CDATA[VV5]]> <![CDATA[V 00 →V 02 →V 22 →V 26 →V 76 →V 77 →V 77 →V 76 →V 26 →V 22 →V 02 →V 00 > <![CDATA[VV6]]> <![CDATA[V 00 →V 20 →V 22 →V 32 →V 37 →V 77 →V 77 →V 37 →V 32 →V 22 →V 20 →V 00 > <![CDATA[VV7]]> <![CDATA[V 00 →V 02 →V 32 →V 33 →V 73 →V 77 →V 77 →V 73 →V 33 →V 32 →V 02 →V 00 > <![CDATA[VV8]]> <![CDATA[V 00 →V 10 →V 13 →V 33 →V 37 →V 77 →V 77 →V 37 →V 33 →V 13 →V 10 →V 00 > <![CDATA[VV9]]> <![CDATA[V 00 →V 01 →V 11 →V 13 →V 73 →V 77 →V 77 →V 73 →V 13 →V 11 →V 01 →V 00 > <![CDATA[VV 10 > <![CDATA[V 00 →V 10 →V 11 →V 51 →V 57 →V 77 →V 77 →V 57 →V 51 →V 11 →V 10 →V 00 > <![CDATA[VV 11 > <![CDATA[V 01 →V 01 →V 51 →V 55 →V 75 →V 77 →V 77 →V 75 →V 55 →V 51 →V 01 →V 00 > <![CDATA[VV 12 > <![CDATA[V 00 →V 40 →V 45 →V 55 →V 57 →V 77 →V 77 →V 57 →V 55 →V 45 →V 40 →V 00 >

[0054] The beneficial effects of adopting the above technical solutions are as follows:

[0055] The present invention provides a dual three-phase motor model predictive current control method with a fixed switching frequency. The advantages of the present invention are as follows: Compared with the traditional dual three-phase permanent magnet motor model predictive current control. The present invention synthesizes virtual voltage vectors, reduces harmonic currents; eliminates the weight coefficients of harmonic terms in the cost function. Designs a process for selecting the optimal voltage vector, reducing the computational burden. Introduces a duty cycle modulation technique to adjust the amplitude of the optimal voltage vector through zero vectors, reducing the error between the optimal voltage vector and the reference voltage vector, and further reducing harmonic currents. Designs the zero vectors at the center and both sides of the control period, fixing the switching frequency. In addition, since the switching sequence of the designed virtual voltage vector is a standard PWM wave, no secondary correction of the switching sequence is required after inserting the zero vectors. Brief Description of the Drawings

[0056] Figure 1 It is the topology diagram of the dual three-phase drive system provided by the embodiment of the present invention;

[0057] Figure 2 It is the voltage vector distribution diagram in the α-β space provided by the embodiment of the present invention;

[0058] Figure 3 It is the voltage vector distribution diagram in the x-y space provided by the embodiment of the present invention;

[0059] Figure 4 It is the schematic diagram of the synthesized vector of the virtual voltage vector provided by the embodiment of the present invention;

[0060] Figure 5 It is the spatial distribution diagram of the virtual voltage vector provided by the embodiment of the present invention;

[0061] Figure 6 It is the schematic diagram of the selection of the optimal voltage vector provided by the embodiment of the present invention;

[0062] Figure 7 It is the selection process diagram of the optimal voltage vector provided by the embodiment of the present invention;

[0063] Figure 8 It is the comparison diagram of the virtual voltage vector PWM wave switching sequences of the traditional method and the present method provided by the embodiment of the present invention;

[0064] Among them, Figure (a) - the comparison diagram of the virtual voltage vector PWM wave switching sequence of the traditional method, and Figure (b) - the comparison diagram of the virtual voltage vector PWM wave switching sequence of the present method;

[0065] Figure 9 This is the switching sequence diagram of the PWM wave combining the virtual voltage vectors and zero vectors of the traditional method and this method provided by the embodiments of the present invention;

[0066] Among them, Figure (a) is the switching sequence diagram of the PWM wave combining the virtual voltage vectors and zero vectors of the traditional method, and Figure (b) is the switching sequence diagram of the PWM wave combining the virtual voltage vectors and zero vectors of this method; Specific embodiments

[0067] The following combines the drawings and embodiments to further describe in detail the specific embodiments of the present invention. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0068] The rated power of the motor used in this embodiment is 28 kW, the rated speed is 3000 r / min, the number of pole pairs is 4, and the DC bus voltage is 340 V. As Figure 1 shown is the two-level dual three-phase permanent magnet synchronous motor drive topology.

[0069] A dual three-phase motor model predictive current control method with a fixed switching frequency specifically includes the following steps:

[0070] Step 1: According to the space decoupling matrix, map the 64 switching states of the six-phase voltage source inverter to the α-β space containing electromechanical energy conversion and the x-y space containing only harmonic components, and obtain the voltage vector distribution of the dual three-phase permanent magnet synchronous motor.

[0071] In the dual three-phase permanent magnet synchronous motor system, according to the VSD transformation, the physical variables in the natural coordinate system can be mapped to the mutually orthogonal α-β space, x-y space, and o1-o2 space. However, the research object of the present invention is a neutral point isolated dual three-phase motor, and there is no o1-o2 space. Among them, the voltage vectors in the α-β space will generate electromagnetic torque during the operation of the motor, while the voltage vectors in the x-y space will generate harmonics and will not generate electromagnetic torque, and the voltage distributions in the α-β space and the x-y space are calculated according to the following formula:

[0072]

[0073] Among them, a = e j30° ; s represents the switching function of the inverter, s i = 1 represents that the upper bridge arm is conducting and the lower bridge arm is off, s i = 0 represents that the upper bridge arm is off and the lower bridge arm is conducting, i represents the A, B, C, U, V, W phases of the inverter; U dc represents the DC bus voltage of the inverter; V αβAnd Vxy are the amplitudes in the α-β space and the x-y space; The six-phase voltage inverter generates 64 voltage vectors, including 60 effective voltage vectors and 4 zero vectors. The 64 basic voltage vectors are divided into four groups according to different amplitudes: large voltage vectors, medium-large voltage vectors, medium-small voltage vectors, and small voltage vectors. The amplitudes are: 0.644U dc , 0.471U dc , 0.331U dc and 0.172U dc , as shown in Figure 2 (α-β space), Figure 3 (x-y space).

[0074] Step 2: Discretize the mathematical model of the dual three-phase permanent magnet synchronous motor through the forward Euler formula to obtain the prediction model of the dual three-phase permanent magnet synchronous motor. Since the model prediction technology optimizes the discretized model, it is derived using the forward Euler formula.

[0075] The mathematical model of the dual three-phase permanent magnet synchronous motor is shown as follows:[[]]

[0076]

[0077]

[0078] where u d and u q are the voltages on the d and q axes; i d and i q are the currents on the d and q axes; u x and u y are the voltages on the x and y axes; i x and i y are the currents on the x and y axes; R s is the stator resistance; ω r is the electrical angular velocity; L d and L q are the inductances on the d and q axes; L ls represents the leakage inductance; ψ f represents the permanent magnet flux linkage;

[0079] In principle, the model prediction method is an optimization method based on the discrete model, and is derived using the forward Euler method:[[]]

[0080]

[0081] where x is a variable; T s represents the sampling period, k represents the k-th sampling period, k + 1 represents the (k + 1)-th sampling period, and the discrete expression of the dual three-phase permanent magnet synchronous motor is expressed as:[[]]

[0082]

[0083] where u d (k) and u q (k) are the voltages on the d- and q-axes at the k-th moment; i d (k) and i q (k) are the currents on the d- and q-axes at the k-th moment;

[0084] The current predictions on the d-axis and q-axis at the k+1 moment are:

[0085]

[0086] To solve the problem of deteriorating predicted current performance caused by digital processing, two-step prediction is used to compensate for the calculation delay, and the current prediction at the instantaneous k+2 is:

[0087]

[0088] Step 3: As Figure 4 shown, to suppress harmonic currents, two large vectors and two medium-small vectors are selected to synthesize virtual voltage vectors, and the virtual voltage vectors are synthesized in sequence according to the duty cycles of 0.432, 0.432, 0.068, and 0.068, and a total of 12 virtual voltage vectors are synthesized;

[0089] The voltage amplitude of the virtual voltage vector in the x-y subspace is zero, and the expression is as follows:

[0090]

[0091] In the formula, 0.173U dc , 0.333U dc are the voltage amplitudes of the large voltage vector and the medium-small voltage vector in the x-y space respectively, where η1, η2, η3, and η4 are the duty cycles of the voltage vectors V 04 , V 44 , V 45 , V 75 respectively, where:

[0092] η1 + η2 + η3 + η4 = 1 (9)

[0093] Substituting formula (9) into formula (8) gives the duty cycle as:

[0094]

[0095] In the α-β space, the amplitude of the virtual voltage vector is obtained according to the volt-second balance principle:

[0096]

[0097] |VV1|αβ is the amplitude of the virtual voltage vector in the α-β space;

[0098] Similarly, 12 virtual voltage vectors with the same amplitude and a phase difference of π / 6 are synthesized. The distribution of the virtual voltage vectors is as Figure 5 shown.

[0099] Step 4: Since the virtual voltage vector is introduced in Step 3, the evaluation function only needs to consider the dq space and does not need to consider the xy space. Since the virtual voltage vector is introduced in Step 3 to suppress the harmonic current, the harmonic term coefficient can be not considered when designing the value function. The weight coefficient is cleverly eliminated, and the designed value function is as shown in Equation (12). Design the value function G without the weight coefficient as follows:

[0100]

[0101] where, i d * (k) and i q * (k) are the given values of the d-axis and q-axis currents respectively; i d (k + 2) and i q (k + 2) are the predicted estimated values of the currents at the (k + 2)-th moment on the d-axis and q-axis respectively.

[0102] Step 5: Design a selection process for the optimal voltage vector to reduce the candidate voltage vectors.

[0103] Figure 6 shows a schematic diagram of selecting the optimal voltage vector, Figure 7 shows the flowchart of selecting the optimal voltage vector.

[0104] Step 5.1: First, perform predictive calculations on the virtual voltage vectors VV4, VV8, VV 12 to obtain the value functions of the three, namely G(VV4), G(VV8), and G(VV 12 ). Compare the value function values of the three, and select the one with the largest value function among the three and denote it as G max , and denote the minimum value function as G min ;

[0105] Step 5.2: Define the voltage vector with the largest value function G max as the worst voltage vector. Due to the voltage vector symmetry of the dual three-phase permanent magnet synchronous motor and the definition of the value function, this means that the reference voltage vector is far from the worst voltage vector, or rather, the reference voltage vector is within a 1 / 3 circle range in the opposite direction of the worst voltage vector. G max determines the range of the reference voltage vector, and the voltage vectors at both ends of the range have a value function of G min, which means the reference voltage vector is closer to the voltage vector with G min , that is, the range of the reference voltage vector is further reduced to 1 / 6 circle. Finally, compare the value functions of the voltage vectors within the 1 / 6 circle range, and the one with the minimum value function is the optimal voltage vector.

[0106] For example: If the position of the reference voltage vector V ref is located at the position in Figure 6 , in the first step, compare the values of G(VV4), G(VV8) and G(VV 12 ), and it can be obtained that G max = G(VV4), then define the worst voltage vector as VV4. As Figure 6 shown, due to the properties of the value function and the voltage vector symmetry of the dual three-phase permanent magnet synchronous motor, this means that the reference voltage vector is far from VV4, or in the range of the opposite direction of the worst voltage vector, that is, in the blue shaded range from VV 12 rotating clockwise to VV8. At this time, calculate G min = G(VV8) at the same time, indicating that the reference voltage vector is far from VV 12 and close to VV8, then narrow the range to the blue shaded range from VV 10 rotating clockwise to VV8. Then, at this time, compare the value functions of VV8, VV9, VV 10 , and the voltage vector with the minimum value function is the optimal voltage vector.

[0107] Step 6: In each control period, determine the dwell time of the optimal voltage vector through duty ratio modulation technology;

[0108] In the traditional model predictive current scheme, only one voltage vector is adopted within one control period, which leads to a large error between the reference voltage vector and the optimal voltage vector, and will cause larger harmonic currents, especially under low sampling frequency conditions. Although virtual voltage vectors are used as the input control set, they still cannot completely eliminate the tracking error. In most cases, it is not necessary to use virtual voltage vectors throughout the control period to meet the required control requirements. The corresponding duty ratios of the selected virtual voltage vectors should be estimated during each control interval to reduce the actual current from accurately tracking the reference value. Therefore, this paper proposes a method combining virtual voltage vectors and zero voltage vectors to adjust the amplitude of the optimal voltage vector and further reduce the tracking error.

[0109] The duty ratio of the optimal voltage vector is:

[0110]

[0111]

[0112] If d is the duty cycle of the optimal voltage vector, then the duty cycle of the zero vector is (1 - d)T s , T s is the control period.

[0113] Among them, the meanings of m and n in Equation (14) are as follows:

[0114]

[0115]

[0116] Step 7: Insert the zero vectors V 00 and V 77 respectively on both sides and in the middle of the switching sequence of the virtual voltage vector, that is, the acting order of the voltage vectors, to achieve the purpose of a fixed switching frequency.

[0117] Step 7.1 The virtual voltage vector is synthesized from four voltage vectors, namely two large voltage vectors and medium and small voltage vectors. If the optimal voltage vector is one of the virtual voltage vectors VV1, VV4, VV5, VV8, VV9 and VV 12 If it is one of them, the switching sequence is designed such that within the first half of the control period, the four voltage vectors act in sequence in the clockwise direction according to their positions in Figure 2 , and in the second half of the period, the four voltage vectors act in sequence in the counterclockwise direction according to their positions in Figure 2 (taking the VV1 switching sequence in (b) as an example). If the optimal voltage vector is one of the virtual voltage vectors VV2, VV3, VV6, VV7, VV Figure 8 , VV 10 , VV 11 If it is one of them, the switching sequence is designed such that within the first half of the control period, the four voltage vectors act in sequence in the counterclockwise direction according to their positions in Figure 2 , and in the second half of the period, the four voltage vectors act in sequence in the clockwise direction according to their positions in Figure 2 .

[0118] Step 7.2 After obtaining the switching sequence of the virtual voltage vector, insert the zero vectors V 00 and V 77 respectively on both sides and in the middle of the switching sequence of the virtual voltage vector. As shown in Figure 9 (b), after inserting the zero vectors, the switching devices of the six phases are turned on and off once within a control period, achieving a fixed switching frequency and forming a standard PWM wave.

[0119] The following gives all the switching sequence designs with a fixed switching frequency of the present invention:

[0120] Table 1 Switching Sequence Table

[0121] Optimal voltage vector Switching sequence with fixed switching frequency <![CDATA[VV1]]> <![CDATA[V 00 →V 04 →V 44 →V 45 →V 75 →V 77 →V 77 →V 75 →V 45 →V 44 →V 04 →V 00 > <![CDATA[VV2]]> <![CDATA[V 00 →V 40 →V 44 →V 64 →V 67 →V 77 →V 77 →V 67 →V 64 →V 44 →V 40 →V 00 > <![CDATA[VV3]]> <![CDATA[V 00 →V 04 →V 64 →V 66 →V 76 →V 77 →V 77 →V 76 →V 66 →V 64 →V 04 →V 00 > <![CDATA[VV4]]> <![CDATA[V 00 →V 20 →V 26 →V 66 →V 67 →V 77 →V 77 →V 67 →V 66 →V 26 →V 20 →V 00 > <![CDATA[VV5]]> <![CDATA[V 00 →V 02 →V 22 →V 26 →V 76 →V 77 →V 77 →V 76 →V 26 →V 22 →V 02 →V 00 > <![CDATA[VV6]]> <![CDATA[V 00 →V 20 →V 22 →V 32 →V 37 →V 77 →V 77 →V 37 →V 32 →V 22 →V 20 →V 00 > <![CDATA[VV7]]> <![CDATA[V 00 →V 02 →V 32 →V 33 →V 73 →V 77 →V 77 →V 73 →V 33 →V 32 →V 02 →V 00 > <![CDATA[VV8]]> <![CDATA[V 00 →V 10 →V 13 →V 33 →V 37 →V 77 →V 77 →V 37 →V 33 →V 13 →V 10 →V 00 > <![CDATA[VV9]]> <![CDATA[V 00 →V 01 →V 11 →V 13 →V 73 →V 77 →V 77 →V 73 →V 13 →V 11 →V 01 →V 00 > <![CDATA[VV 10 > <![CDATA[V 00 →V 10 →V 11 →V 51 →V 57 →V 77 →V 77 →V 57 →V 51 →V 11 →V 10 →V 00 > <![CDATA[VV 11 > <![CDATA[V 01 →V 01 →V 51 →V 55 →V 75 →V 77 →V 77 →V 75 →V 55 →V 51 →V 01 →V 00 > <![CDATA[VV 12 > <![CDATA[V 00 →V 40 →V 45 →V 55 →V 57 →V 77 →V 77 →V 57 →V 55 →V 45 →V 40 →V 00 >

[0122] For a dual-three-phase permanent magnet synchronous motor, half of the virtual voltage vectors synthesized by using the medium-large vectors and large vectors in the large vector sum cannot implement standard PWM waves. For example: Figure 8 As shown in (a), for V 64 and V 46 the switching sequence of the synthesized virtual voltage vector is not a standard PWM wave. This results in that after inserting zero vectors, Figure 9 the switching device of the V phase in (a) is turned on and off 3 times, which leads to a variable switching frequency of the system. In order to generate such a PWM wave, a piecewise function needs to be used for the carrier signal within one sampling period, which increases the workload in software design. Further research reveals that whether it is the virtual voltage vector synthesized simply by using the medium-large vectors and large vectors or the virtual voltage vector synthesized by using three adjacent large voltage vectors, the virtual voltage vector cannot form a completely standard PWM wave. The virtual voltage vector synthesized by the present invention can solve this problem, Figure 8 as shown in (b). Therefore, the virtual voltage vector of the present invention can also generate a standard PWM wave after inserting zero vectors. For example: the optimal voltage vector is VV1. The switching sequence generated by the combination of VV1 and zero vectors is as Figure 9 shown in (b). This switching sequence is a standard PWM wave, and the corresponding harmonic content of the symmetric PWM wave is less. The proposed switching sequence is V 00 →V 04 →V 44 →V 45 →V 75 →V 77 →V 77 →V 75 →V 45 →V 44 →V 04 →V 00 . A virtual voltage vector capable of generating a standard PWM wave is directly designed, so that it can be directly combined with zero vectors for a fixed switching frequency without the need for correction of the two switching sequences.

[0123] The above description is only the preferred embodiment of the present disclosure and the explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A method for model predictive current control of a dual three-phase motor with a fixed switching frequency, characterized in that It includes the following steps: Step 1: According to the space decoupling matrix, map the 64 switching states of the six-phase voltage source inverter to the α-β space containing electromechanical energy conversion and the x-y space with only harmonic components, and obtain the voltage vector distribution of the dual three-phase permanent magnet synchronous motor; Step 2: Discretize the mathematical model of the dual three-phase permanent magnet synchronous motor through the forward Euler formula to obtain the prediction model of the dual three-phase permanent magnet synchronous motor; Step 3: Select two large vectors and two medium and small vectors to synthesize virtual voltage vectors, and synthesize virtual voltage vectors in turn according to the duty ratios of 0.432, 0.432, 0.068, and 0.068, and a total of 12 virtual voltage vectors are synthesized; Step 4: Design a value function G without weight coefficients as follows: where, i d * (k) and i q * (k) are the reference values of the d-axis and q-axis currents respectively; i d (k + 2) and i q (k + 2) are the predicted estimated values of the currents at the (k + 2)-th moment on the d-axis and q-axis respectively; Step 5: Design a selection process for the optimal voltage vector to reduce the candidate voltage vectors; Step 6: In each control cycle, determine the dwell time of the optimal voltage vector through the duty ratio modulation technique; Step 7: A switching sequence with a fixed switching frequency is designed. Zero vectors V 00 and V 77 are inserted respectively on both sides and in the middle of the switching sequence of the virtual voltage vector, i.e., the acting order of the voltage vectors, so as to achieve the purpose of a fixed switching frequency.

2. A method for predicting current control of a dual-three-phase motor with a fixed switching frequency according to claim 1, characterized in that The voltage vector distribution described in Step 1, that is, the voltage distributions in the α-β space and the x-y space are shown in the following formula: Among them, a = e j30° ; s represents the switching function of the inverter, s i = 1 represents that the upper bridge arm is conducting and the lower bridge arm is off, s i = 0 represents that the upper bridge arm is off and the lower bridge arm is conducting, i represents the A, B, C, U, V, W phases of the inverter; U dc represents the DC bus voltage of the inverter; V αβ and Vxy are the amplitudes in the α-β space and x-y space; the six-phase voltage inverter generates 64 voltage vectors, including 60 effective voltage vectors and 4 zero vectors. The 64 basic voltage vectors are divided into four groups according to different amplitudes: large voltage vectors, medium-large voltage vectors, medium-small voltage vectors, and small voltage vectors. The amplitudes are: 0.644U dc , 0.471U dc , 0.331U dc and 0.172U dc .

3. A method for model predictive current control of a dual-three-phase motor with a fixed switching frequency according to claim 1, characterized in that, The mathematical model of the dual three-phase permanent magnet synchronous motor described in Step 2 is shown in the following formula: where u d and u q are the voltages on the d- and q-axes; i d and i q are the currents on the d- and q-axes; u x and u y are the voltages on the x- and y-axes; i x and i y are the currents on the x- and y-axes; R s is the stator resistance; ω r is the electrical angular velocity; L d and L q are the inductances on the d- and q-axes; L ls represents the leakage inductance; ψ f represents the permanent magnet flux linkage; Derivation is carried out using the forward Euler method: where x is a variable; T s represents the sampling period, k represents the k-th moment, k + 1 represents the (k + 1)-th moment, and the discrete expression of the dual three-phase permanent magnet synchronous motor is expressed as: where u d (k) and u q (k) are the voltages on the d- and q-axes at the k-th moment; i d (k) and i q (k) are the currents on the d- and q-axes at the k-th moment; The current predictions on the d-axis and q-axis at the k + 1 moment are: Using two-step prediction to compensate for the calculation delay, the current prediction at the instantaneous k + 2 is:

4. A method for model predictive current control of a dual-three-phase motor with a fixed switching frequency according to claim 1, characterized in that The voltage amplitude of the virtual voltage vector described in Step 3 on the x-y subspace is zero, and the expression is as follows: Where U dc represents the DC bus voltage of the inverter, and 0.173U dc , 0.333U dc are the voltage amplitudes of the large voltage vector and the medium and small voltage vectors in the x-y space respectively, where η1, η2, η3, and η4 are the duty cycles of the four voltage vectors V 04 , V 44 , V 45 , V 75 respectively, where: η1 + η2 + η3 + η4 = 1 (9) Substitute formula (9) into formula (8) to obtain the duty ratio: In the α-β space, the amplitude of the virtual voltage vector is obtained according to the volt-second balance principle: |VV1| αβ is the magnitude of the virtual voltage vector in the α-β space.

5. A method for predicting current control of a dual-three-phase motor with a fixed switching frequency according to claim 1, characterized in that The specific content of Step 5 includes the following steps: Step 5.1: First, perform predictive calculations on the virtual voltage vectors VV4, VV8, VV 12 to obtain their value functions, namely G(VV4), G(VV8), and G(VV 12 ). Compare the values of these three value functions and select the one with the largest value function among them, denoted as G max , and denote the minimum value function as G min ; Step 5.2: Define the voltage vector with the largest value function G max as the worst voltage vector. The reference voltage vector is within a 1 / 3 circle range in the opposite direction of the worst voltage vector, G max determines the range of the reference voltage vector, and there is a voltage vector with a value function of G at both ends within the range min . Compare the value functions of the voltage vectors within the 1 / 6 circle range, and the one with the smallest value function is the optimal voltage vector.

6. A method for predicting current control of a dual-three-phase motor with a fixed switching frequency according to claim 1, characterized in that The duty ratio of the optimal voltage vector in Step 6 is: G is the value function, u d and u q are the voltages on the d- and q-axes, L d and L q are the inductances on the d- and q-axes, R s is the stator resistance, ω r is the electrical angular velocity, ψ f is the permanent magnet flux linkage, d is the duty ratio of the optimal voltage vector, and the duty ratio of the zero vector is (1 - d)T s , T s is the control period; Among them, the meanings of m and n in formula (14) are as follows:

7. A method for model predictive current control of a dual three-phase motor with a fixed switching frequency according to claim 1, characterized in that, The switching sequence of the fixed switching frequency described in Step 7 is shown in Table 1; Table 1 Switching Sequence Table Step 7.1: The virtual voltage vector is synthesized from four voltage vectors, namely two large voltage vectors and medium and small voltage vectors. If the optimal voltage vector is one of the virtual voltage vectors VV1, VV4, VV5, VV8, VV9, and VV 12 Among them, the switching sequence is designed such that within the first half of the control period, the four voltage vectors act in sequence in the clockwise direction, and in the second half of the period, the four voltage vectors act in sequence in the counterclockwise direction. If the optimal voltage vector is one of the virtual voltage vectors VV2, VV3, VV6, VV7, VV 10 , VV 11 Among them, the switching sequence is designed such that within the first half of the control period, the four voltage vectors act in sequence in the counterclockwise direction, and in the second half of the period, the four voltage vectors act in sequence in the clockwise direction; Step 7.2: After obtaining the switching sequence of the virtual voltage vector, insert the zero vectors V 00 and V 77 respectively on both sides and in the middle of the switching sequence of the virtual voltage vector. After inserting the zero vectors, the switching devices of the six phases are turned on and off once within one control period, achieving a fixed switching frequency and forming a standard PWM wave.

Images